Not acheiving the expected hardness for T23 regardless of cooling rate

[BWmatleng]

I have a T23 tube with a reported analysis of:
carbon 0.07 (ASTM required 0.04 - 0.10)
manganese 0.45 (0.10 - 0.60)
silicon 0.32 (0.50 max)
nickel 0.11 (not listed)
chromium 2.08 (1.90 - 2.60)
tungsten 1.61 (1.45 - 1.75)
molybdenum 0.12 (0.05 - 0.30)
copper 0.15 (not listed)
titanium 0.001 (not listed)
vanadium 0.22 (0.20 - 0.30)
niobium 0.037 (0.02 - 0.08)
aluminum 0.003 (0.03 max)
boron 0.002 (0.0005 - 0.005)
nitrogen 0.013 (0.03 max)

2 different tube manufacturers estimate that heating this material at 1940F with air cool should result in 300-320 BHN. With a reasonable quench from the same temp the resulting hardness should be 320-350 BHN.

While evaluating a new heat treater, a third tube manufacturer found that the hardness was not met after normalizing in a continuous roller hearth furnace. The resulting hardness was 156/179 BHN.

Further testing by the manufacturer showed:
Normalize in a batch furnace (a different heat treat vendor) resulted in 207/217 BHN.
Normalize in a lab furnace resulted in 217/229 BHN, with a microstructure of blocky ferrite and fine, unresolvable pearlite.
Air quenching with fans resulted in 235/262 BHN, with a microstructure of blocky ferrite, but this time with fine pearlite and possibly some small amount of bainite.
Water quenching resulted in 277/311 BHN, with a microstructure that was essentially bainitic.

In every case the hardness appears to be 70-80 brinell points below what the curves predict.

We found:
A normalize in a lab furnace resulted in 228 BHN with a microstructure of blocky ferrite, unresolvable pearlite, some amount bainite.
A normalize in a lab furnace using a fan to accelerate air cooling resulted in 216/222 BHN with a microstructure of what looked like ferrite and bainite.
A water quench after heating in a lab furnace resulted in 258 BHN, but the structure still looked like ferrite and bainite.
Heating in a lab furnace with a water quench of P23 pipe resulted in 250/253 BHN with a microstructure of bainite.

We obtained a different heat of T23, from a different manufacturer with a reported analysis of:
carbon 0.07
manganese 0.49
silicon 0.24
nickel 0.13
chromium 2.09
tungsten 1.70
molybdenum 0.17
copper 0.10
titanium --- (not listed)
vanadium 0.221
niobium 0.031
aluminum 0.012
boron 0.002
nitrogen 0.0076

The MTR states the hardness after N&T as 179/191 BHN.

We normalized the sample in a lab furnace with air cooling (no acceleration of cooling) and the resulting hardness was 216/234 BHN with a microstructure that looked like ferrite and bainite.

Does anyone have an idea of why the required hardness that is predicted by industry accepted curves is not being acheived? Any ideas on why we can't get bainite, as expected?

 

[metengr]

Does anyone have an idea of why the required hardness that is predicted by industry accepted curves is not being acheived? Any ideas on why we can't get bainite, as expected?

What industry accepted curves? Have you contacted V&M or Sumitomo directly? I have dealt with both and they will work directly with end users like myself.

 

[BWmatleng]

I have been using the curves from V&M to determine the expected hardness and microstructure. If we can't come up with a reasonable theory for the differences, we are probably going to send the question to Benteler or V&M for input.

 

[metengr]

BWmatleng;
I had visited Benteler in Germany last year on a large tube order for one of our Generating Stations. I do not recall that Benteler handled Grade T23 tubing because they do not have VOD or vacuum degassing handling processes for this material. They have EAF air melt capability and strand casting. I recall they make only up to 9 %Cr-Mo steels along with a variety of carbon and low alloy.

I would suggest you contact either V&M or Sumitomo directly. By the way, do you work for B&W out of the Barberton, Ohio office? I know most of the mets there.

 

[BWmatleng]

Yes, I work for B&W in Barberton.

I mention Benteler because we received a sample of T23 tubing from them. It was the last sample described above. This isn't my project, by the way, I was looking for any information that I could pass along to the engineer that is working on the issue. It isn't often that the cooling rate used misses the nose of the curve but doesn't result in the predicted hardness or microstructure.

 

[metengr]

BWmatleng;
The hardness data reported for four steel samples supplied as part of the original code case range from 183 Hv to 202 Hv.

The SA 213 tube specification calls for a maximum hardness value of 230 Hv based on the above review of T23 steel data that was submitted for the code case.

The CCT that was submitted for HCM2S (prior to becoming Grade T23) shows that for a cooling rate higher than 18 deg C per minute, you should achieve 100% bainite (where the hardness is greater than 285 Hv). By the way, at cooling rates lower than 18 deg C/min, I see both ferrite and bainite in the submitted micrographs and the hardness is as above (180-203 Hv).

I am sure if you are out of the Barberton office you know Jim Tanzosh, he was involved with the original code case submittal and should have the data package that I have and more. This thread may help as well;

thread330-219065

 

[metengr]

BWmatleng;
One other item in reviewing your chemical analysis results, the boron content in your steel heats above seems too low. You really should have a minimum of 0.003% and closer to 0.004% by mass content of boron in these heats for improved hardenability. This may be the key.

 

[BWmatleng]

Metengr;
Thank you for the lead on Boron. We are going to look at that next. When I passed your information on to Hainsworth, he asked who posted the information. I sat in the ASME Section II Committee meeting in Boston and Vancouver this year and have probably met you.

 

[deadrange]

Metengr:
I am intrigued by your assessment that additional boron needs to be added to the steel. Traditionally, boron treated grades 0.001 wt % is sufficient to get the improved hardenability. What is the rationale behind the need for 0.003 to 0.004 wt % boron?

 

[davefitz]

As I recall from similar issues with P91, the time and temp for the normalizing will also affect grain size and thus hardness, regardless of quench rate. Too high a normalizing temp will result in large grain sizes and low hardness. Calibrate your furnace thermocouples and ensure they are shielded from radiative cooling to some cooler surface.

 

[Welder4956]

metengr

What happened with T23 at the November 2008 TG-CSEF and SCII meetings? I'm hearing rumblings there was a move to withdraw the code case due to concerns over low hardness values after normalizing and the chemistry range being too broad for some elements. Can you shed any light on this?

 

[metengr]

Welder4956;
From the recent ASME B&PV Code meeting in November, the SG on strength of ferrous alloys and the TG-CSEF are re-evaluating Code Case 2199 to decide on what needs to be revised regarding the original chemical composition provided by Sumitomo. The Code Case was close to being revoked at this meeting. However, cooler heads had prevailed and it would appear that resolution to this problem is expected at the February, 2009 meeting.

For users that have Grade T23 installed in their boiler or HRSG, a word of caution is to check on the source of the mill that supplied the Grade T23 boiler tubing. If this tubing was supplied by mills other than V-M or Sumitomo you may be at risk for having creep rupture failures sooner rather than later.

 

[Welder4956]

Hmmm...what about tubes from JFE Corp.?

 

[metengr]

Same applies to others. Check the material..............

 

[deadrange]

metengr,

Is there going to be an sort of notification from ASME about grade T23?

 

[metengr]

deadrange;
In my mind, we have two immediate issues to deal with; the first issue is users that have Grade T23 boiler tubing that was supplied by mills other than Sumitomo or V&M. Since this material has already been installed, this is no longer an ASME B&PV Code matter and falls under the National Board domain and perhaps EPRI. This has been dealt with by an alert to industry users from EPRI. My own recommendation to users is to determine the source of Grade T23, and work with the supplier of the tubing to evaluate creep behavior.

The second issue is what to do with the existing Code Case. This I believe will be handled during the upcoming ASME B&PV Code meeting scheduled in February, 2009. In the meantime, I would not place any order for Grade T23 until the original Code Case has been revised. I am guessing that the chemical composition ranges will be significantly tightened with respect to boron, and titanium may be addressed as a separate issue.

The summary I provided above is my own interpretation and consider it an alert for users that visit this forum.

 

[deadrange]

metengr,

Is there a link for the EPRI warning? I have quickly scanned their website and didn't see anything. Granted I am not a member of EPRI, so I am not sure I would be able to see this alert.

 

[metengr]

deadrange;
No, this alert only applies to members of EPRI Generation Programs (87). However, I summarized in so many words what was in the alert to you and others with the exception of company names.

What else do you need to know about this?